The Development of Crash Modification Factors: Highway Safety Statistical Paper Synthesis

Monday, January 18, 2021 - 20:15

The transportation-engineering community is transforming by integrating quantitative methods into the task development process. This report identifies opportunities to better understand the relationships between road safety and factors that affect traffic-crash occurrence and severity. In this report, current statistical-analysis methods and data sources used in road-safety research are compared with alternative methods and data sources. Causal inference methods are compared to observational before–after methods to develop safety-effect estimates of centerline and edgeline rumble strips. Regression trees and Random ForestsTM are compared to count regression methods to predict crash frequencies on freeways. Road-safety performance estimates using the Crash Outcomes Data Evaluation System are also discussed, with a focus on opportunities to link hospital and crash data to understand the relationship between crashes and site-specific contributing factors. Methods to account for underreporting in crash-frequency models are also described.  This report is available from the US Federal Highway Administration at


Thank you to our Premier Sponsors